The big political news of the morning, going into the Tea Party debate tonight, is that the Minnesota Meh, Tim Pawlenty, has endorsed Mitt Romney.
Endorsements are tricky things to understand in presidential nomination politics, because they can mean different things. Let's see...
Collectively, high-profile endorsements are a way for party actors to signal to each other their preferences, which helps them to coordinate their collective choice -- or, in the case of factional battles over the nomination, to coordinate a faction. In that sense, endorsements are quite important.
But individual endorsements may or may not play into that, because endorsers may be acting as part of the party or on behalf of individual, perhaps idiosyncratic reasons. Now, some self-interest may line up with party coordination. For example, if Pawlenty's goal is to get a cabinet post or the VP nod, then he has a strong incentive to figure out who is likely to win the nomination -- and he thus sends an informed signal to others about how the contest is going. On the other hand, it's not unusual for presidential candidates to wind up holding grudges against each other and acting on that after they leave the race.
Note that in all of this endorsements can be both a cause of future consolidation behind a candidate and an effect of previous candidate success.
I should mention too that endorsements can bring resources (money, volunteer time) with them, although it's best to be careful about assuming that a failed presidential candidate controls, or even has a strong influence on, his former supporters. Might be true, might not. Press reports sometimes are a bit too quick to make those connections; indeed, exaggerated reports of resources that supposedly came with endorsements may account for the idea that endorsements don't matter, which I also hear frequently. So I'd be cautious about that sort of effect -- although surely it can exist, to some extent.
So all in all endorsements are certainly worth paying a lot of attention to, but they tell us a lot in the aggregate that isn't necessarily true of individual cases.